Conventional speech recognition systems on board an automobile facilitate drivers and/or passengers to control various vehicle functions through voice commands. Typically, one or more microphones can be placed throughout the cabin a the automobile and coupled to a speech processor to process user voice commands using acoustic and language modeling techniques. Voice pattern data is typically stored in a memory in an on-board speech recognition system and is retrieved from the memory for recognizing user voice commands.
One problem associated with the conventional on-board speech recognition system is that it is limited in recognizing voice commands provided by user with a dialect. A dialect typically refers to a variety of a single language that is a characteristic of a particular group of the language's speakers. Different communities within the single language may have different ways of talking the same language that set them apart from others. Those differences may be thought of as dialects—not just accents (the way words are pronounced) but also grammar, vocabulary, syntax and common expressions. Often a group that is somewhat isolated regionally or socially from other groups will develop a characteristic dialect. For example, there are thousands of dialects in China such as Cantonese, Shanghainese, Sichuanese, Fujianese and on and on.
Users with different dialects present a challenge to conventional on-board speech recognition system in that the conventional on-board speech recognition system only maps a user input to a standard voice pattern. That is, the conventional on-board speech recognition system is not equipped to distinguish dialect differences and requires a user to input voice command through a standard language pattern. In the case of Chinese, the conventional on-board speech recognition system typically requires a user to input voice commands through Mandarin. If a user inputs a voice command through his/her dialect, the conventional on-board speech system would typically not respond correctly.